In [1]:
import pandas as pd
import numpy as np
from sklearn import metrics
%matplotlib inline
import matplotlib.pyplot as plt
In [2]:
df = pd.read_csv(r"C:\Users\advait\Downloads\HDFCBANK.csv")
In [3]:
df.head()
Out[3]:
Date Symbol Series Prev Close Open High Low Last Close VWAP Volume Turnover Trades Deliverable Volume %Deliverble
0 2000-01-03 HDFCBANK EQ 157.40 166.00 170.00 166.00 170.00 170.00 169.52 33259 5.638122e+11 NaN NaN NaN
1 2000-01-04 HDFCBANK EQ 170.00 182.00 183.45 171.00 174.00 173.80 174.99 168710 2.952261e+12 NaN NaN NaN
2 2000-01-05 HDFCBANK EQ 173.80 170.00 173.90 165.00 168.00 166.95 169.20 159820 2.704094e+12 NaN NaN NaN
3 2000-01-06 HDFCBANK EQ 166.95 168.00 170.00 165.30 168.95 168.30 168.44 85026 1.432166e+12 NaN NaN NaN
4 2000-01-07 HDFCBANK EQ 168.30 162.15 171.00 162.15 170.75 168.35 166.79 85144 1.420158e+12 NaN NaN NaN

# PRE-PROCESSING

In [4]:
df.drop(['Trades','Deliverable Volume','%Deliverble'],axis=1,inplace=True)
In [5]:
series_shifted = df['Close'].shift()
df['Prev Close'] = series_shifted
df.head()
Out[5]:
Date Symbol Series Prev Close Open High Low Last Close VWAP Volume Turnover
0 2000-01-03 HDFCBANK EQ NaN 166.00 170.00 166.00 170.00 170.00 169.52 33259 5.638122e+11
1 2000-01-04 HDFCBANK EQ 170.00 182.00 183.45 171.00 174.00 173.80 174.99 168710 2.952261e+12
2 2000-01-05 HDFCBANK EQ 173.80 170.00 173.90 165.00 168.00 166.95 169.20 159820 2.704094e+12
3 2000-01-06 HDFCBANK EQ 166.95 168.00 170.00 165.30 168.95 168.30 168.44 85026 1.432166e+12
4 2000-01-07 HDFCBANK EQ 168.30 162.15 171.00 162.15 170.75 168.35 166.79 85144 1.420158e+12
In [6]:
df.dropna(inplace=True)
df[['Date','Symbol','Open','High','Low','Close','Prev Close']]
Out[6]:
Date Symbol Open High Low Close Prev Close
1 2000-01-04 HDFCBANK 182.00 183.45 171.00 173.80 170.00
2 2000-01-05 HDFCBANK 170.00 173.90 165.00 166.95 173.80
3 2000-01-06 HDFCBANK 168.00 170.00 165.30 168.30 166.95
4 2000-01-07 HDFCBANK 162.15 171.00 162.15 168.35 168.30
5 2000-01-10 HDFCBANK 172.90 179.50 165.00 165.90 168.35
... ... ... ... ... ... ... ...
5301 2021-04-26 HDFCBANK 1413.00 1429.00 1402.75 1404.80 1414.15
5302 2021-04-27 HDFCBANK 1407.25 1442.00 1404.80 1438.70 1404.80
5303 2021-04-28 HDFCBANK 1436.25 1479.00 1431.00 1476.80 1438.70
5304 2021-04-29 HDFCBANK 1486.20 1503.65 1461.00 1472.50 1476.80
5305 2021-04-30 HDFCBANK 1445.00 1453.80 1407.50 1412.30 1472.50

5305 rows × 7 columns

In [7]:
df.info()

<class 'pandas.core.frame.DataFrame'>
Index: 5305 entries, 1 to 5305
Data columns (total 12 columns):
 #   Column      Non-Null Count  Dtype  
---  ------      --------------  -----  
 0   Date        5305 non-null   object 
 1   Symbol      5305 non-null   object 
 2   Series      5305 non-null   object 
 3   Prev Close  5305 non-null   float64
 4   Open        5305 non-null   float64
 5   High        5305 non-null   float64
 6   Low         5305 non-null   float64
 7   Last        5305 non-null   float64
 8   Close       5305 non-null   float64
 9   VWAP        5305 non-null   float64
 10  Volume      5305 non-null   int64  
 11  Turnover    5305 non-null   float64
dtypes: float64(8), int64(1), object(3)
memory usage: 538.8+ KB
In [8]:
df.describe()
Out[8]:
Prev Close Open High Low Last Close VWAP Volume Turnover
count 5305.000000 5305.000000 5305.000000 5305.000000 5305.000000 5305.000000 5305.000000 5.305000e+03 5.305000e+03
mean 1007.254053 1007.631385 1020.147163 993.978256 1007.521847 1007.488228 1007.087116 2.102970e+06 2.449297e+14
std 635.710618 635.416358 641.398943 629.459517 635.678685 635.630950 635.437777 3.970321e+06 4.778436e+14
min 163.400000 162.150000 167.900000 157.000000 163.000000 163.400000 161.400000 1.042000e+03 2.291142e+10
25% 479.950000 482.750000 486.950000 473.400000 481.000000 480.950000 480.150000 3.061880e+05 1.642027e+13
50% 935.000000 939.700000 954.000000 922.200000 936.200000 936.450000 937.470000 9.445620e+05 1.150500e+14
75% 1421.150000 1424.000000 1440.000000 1399.000000 1423.000000 1421.150000 1420.780000 2.124352e+06 2.154579e+14
max 2565.800000 2566.000000 2583.300000 2553.700000 2563.000000 2565.800000 2570.700000 1.005650e+08 1.426400e+16
In [9]:
df['Date'] = pd.to_datetime(df.Date)
In [10]:
import plotly.graph_objects as go


fig = go.Figure()
fig.add_trace(go.Scatter(x=df['Date'], y=df['Close'], mode='lines', name='Closing Price'))


fig.update_layout(
    title='Stock Closing Prices Over Time',
    xaxis_title='Date',
    yaxis_title='Closing Price',
    width=1000, 
    height=600
)

C:\Users\advait\anaconda3\lib\site-packages\_plotly_utils\basevalidators.py:105: FutureWarning:

The behavior of DatetimeProperties.to_pydatetime is deprecated, in a future version this will return a Series containing python datetime objects instead of an ndarray. To retain the old behavior, call `np.array` on the result

# MODEL CREATION

In [11]:
x = df[['Open','Prev Close']]
y = df['Close']
In [12]:
from sklearn.model_selection import train_test_split
x_train,x_test,y_train,y_test = train_test_split(x,y,test_size=0.33,random_state =42)
In [13]:
print(x_train.shape,x_test.shape)

(3554, 2) (1751, 2)
In [14]:
from sklearn.linear_model import LinearRegression
from sklearn.metrics import confusion_matrix, accuracy_score

reg = LinearRegression()
reg.fit(x_train,y_train)
Out[14]:
LinearRegression()
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
LinearRegression()
In [15]:
print(reg.coef_)
print(reg.intercept_)

[0.94027433 0.05982184]
-0.11979445269525968
In [16]:
predicted = reg.predict(x_test)
print(x_test)

        Open  Prev Close
812    235.5      234.30
4039  1027.0     1026.70
4798  2250.0     2287.25
3066   536.4      530.40
1865  1105.5     1090.95
...      ...         ...
1698   900.3      890.15
137    250.0      251.25
4632  2121.0     2121.65
5217  1435.0     1441.80
577    226.8      224.75

[1751 rows x 2 columns]
In [17]:
_df = pd.DataFrame({"Actual":y_test, "Predicted":predicted})
print(_df)

       Actual    Predicted
812    233.15   235.331067
4039  1030.45  1026.961022
4798  2237.35  2252.324944
3066   529.35   535.972858
1865  1097.35  1104.616110
...       ...          ...
1698   931.00   899.659593
137    248.35   249.979024
4632  2114.30  2121.123059
5217  1411.35  1435.424993
577    224.95   226.579381

[1751 rows x 2 columns]
In [18]:
reg.score(x_test,y_test)
Out[18]:
0.9991329315822302
In [19]:
import math

print("Mean Absolute Error",metrics.mean_absolute_error(y_test,predicted))
print("Mean Squared Error",metrics.mean_squared_error(y_test,predicted))
print("Root Mean Error",math.sqrt(metrics.mean_squared_error(y_test,predicted)))

Mean Absolute Error 12.059358949887846
Mean Squared Error 349.361136361257
Root Mean Error 18.691204786242565
In [20]:
predicted = reg.predict(x)
print(predicted)

[ 181.1798458   170.12387685  167.8335486  ... 1436.41488831 1485.66080304
 1446.66426683]

### RESULT

In [21]:
fig = go.Figure()


fig.add_trace(go.Scatter(x=list(range(len(y))), y=y, mode='lines', name='Actual'))


fig.add_trace(go.Scatter(x=list(range(len(predicted))), y=predicted, mode='lines', name='Predicted'))


fig.update_layout(
    title='Actual vs Predicted',
    xaxis_title='Index',
    yaxis_title='Value',
    legend=dict(x=0, y=1),
    width=1000,
    height=600
)


fig.show()
In [ ]: